Title Population substructure and signals of divergent adaptive selection despite admixture in the sponge Dendrilla antarctica from shallow waters surrounding the Antarctic Peninsula Authors Leiva, Carlos; Taboada, Sergi; Kenny, Nathan J.; Combosch, David; Giribet, Gonzalo; Jombart, Thibaut; Riesgo, Ana Date Submitted 2019-08-08 1 Population substructure and signals of divergent adaptive selection despite admixture in 2 the sponge Dendrilla antarctica from shallow waters surrounding the Antarctic 3 Peninsula 4 5 Running head: Admixture & adaptation in an Antarctic sponge 6 7 Authors 8 Carlos Leiva1,2*, Sergi Taboada1,3, Nathan J. Kenny1, David Combosch4, Gonzalo Giribet5, 9 Thibaut Jombart6, Ana Riesgo1 10 11 1 Department of Life Sciences, The Natural History Museum of London, Cromwell Road, 12 London SW7 5BD, UK. 13 2 Department of Genetics, Microbiology and Statistics, Faculty of Biology, University of 14 Barcelona, Avinguda Diagonal 643, 08028, Barcelona, Spain. 15 3 Department of Biology, Faculty of Science, Autonomous University of Madrid, Darwin 16 Street, 2, 28049, Madrid, Spain. 17 4 Marine Laboratory, University of Guam, Mangilao, GU, USA. 18 5 Deparment of Organismic and Evolutionary Biology, Harvard University, 26 Oxford Street, 19 02138 Cambridge, USA. 20 6 Department of Infectious Disease Epidemiology, London School of Hygiene and Tropical 21 Medicine, Keepel Street, London WC1E 7HT, UK. 22 *Corresponding author: [email protected] 1 23 Abstract 24 Antarctic shallow-water invertebrates are exceptional candidates to study population genetics 25 and evolution, because of their peculiar evolutionary history and adaptation to extreme 26 habitats that expand and retreat with the ice sheets. Among them, sponges are one of the 27 major components, yet population connectivity of none of their many Antarctic species has 28 been studied. In order to investigate gene flow, local adaptation, and resilience to near-future 29 changes caused by global warming, we sequenced 62 individuals of the sponge Dendrilla 30 antarctica along the Western Antarctic Peninsula (WAP) and the South Shetlands (~ 900 km). 31 We obtained information from 577 ddRADseq-derived SNPs, using RADseq techniques for 32 the first time with shallow-water sponges. In contrast to other studies in sponges, our 389 33 neutral SNPs dataset showed high levels of gene flow, with a subtle substructure driven by the 34 circulation system of the studied area. However, the 140 outlier SNPs under positive selection 35 showed signals of population differentiation, separating the central-southern WAP from the 36 Bransfield Strait area, indicating a divergent selection process in the study area despite 37 panmixia. Fourteen of these outliers were annotated, being mostly involved in immune and 38 stress responses. We suggest that the main selective pressure on D. antarctica might be the 39 difference in the planktonic communities present in central-southern WAP compared to the 40 Bransfield Strait area, ultimately depending on sea-ice control of phytoplankton blooms. Our 41 study unveils an unexpectedly long distance larval dispersal exceptional in Porifera, 42 broadening the use of genome-wide markers within non-model Antarctic organisms. 43 44 Keywords: adaptation, ddRADseq, mitochondrial genome, RNA-Seq, SNPs, South Shetland 45 Islands 2 46 1 | INTRODUCTION 47 The gene flow and phylogeographic patterns of Southern Ocean shallow-water marine 48 invertebrates in general, and sponges in particular, are interesting for a number of reasons. 49 From an evolutionary history perspective, the Southern Ocean provides a unique scenario for 50 studying the impact of drastic environmental shifts on the population dynamics of marine 51 species, with repeated Pliocene–Pleistocene glacial cycles being the major factor in shaping 52 the current diversity and distribution of the Antarctic fauna (Thatje, Hillenbrand, & Larter, 53 2005). The alternation of glacial and interglacial periods might have especially affected 54 shallow-water benthic invertebrates, eliminating most of the available habitat during glacial 55 maxima (Thatje et al., 2005; Rogers, 2007; Allcock & Strugnell, 2012). These dramatic 56 environmental events left characteristic signatures throughout the genome of these shallow- 57 water invertebrates, most of which have only been assessed using traditional mitochondrial [a 58 fragment of the cytochrome c oxidase subunit I gene (COI)] and nuclear (18S and 28S rRNAs 59 genes) markers (e.g. Krabbe, Leese, et al., 2010; Janosik, Mahon, & Halanych, 2011; 60 González-Wevar, David, & Pulin, 2011; González-Wevar, Saucède, et al., 2013). Although 61 shallow-water sponges form massive reefs dominating an important fraction of the available 62 hard substrate in Antarctica (Dayton, 1989), no study has yet addressed the population 63 genetics and connectivity of any of the 397 described sponges from this continent (Downey, 64 Griffiths, Linse, & Janussen, 2012; Riesgo, Taboada, & Avila, 2015). To our knowledge, the 65 only study incorporating analyses of the genetic diversity of an Antarctic sponge was 66 conducted on the deep-sea species Stylocordyla chupachups using microsatellites (Carella, 67 Agell, & Uriz, 2018), but we did not consider it as a population genetics and connectivity 68 study because the authors only focused on the sponge clonal reproduction at a very small 3 69 scale (less than 2 km). 70 The Antarctic Peninsula is currently one of the most rapidly warming regions of the 71 planet (Vaughan, Marshall, et al., 2003). The mean atmospheric temperature rose nearly 3 ºC 72 during the second half of the 20th century (King, 1994; King & Harangozo, 1998; Turner, 73 Colwell, et al., 2005), with profound consequences for ice sheets and glaciers (Cook, Fox, 74 Vaughan, & Ferrigno, 2005). Moreover, the summer temperature of the surface waters 75 adjacent to the WAP increased by more than 1 ºC during the same period (Meredith & King, 76 2005), threatening shallow-water Antarctic species, which are less resilient to temperature 77 increases than species elsewhere (Peck & Conway, 2000), and whose essential biological 78 functions are extremely sensitive to temperature fluctuations (Peck, Webb, & Bailey, 2004). 79 This aspect is especially concerning for sponges, as all of them are sessile organisms known 80 to have lecithotrophic larvae (Maldonado, 2006), which would imply limited dispersal 81 abilities and therefore higher vulnerability (Pascual et al. 2017). However, although in the 82 Southern Ocean the reproductive life history stages appear to have little influence in 83 structuring genetic patterns (Halanych & Mahon, 2018), sponge larvae from other latitudes 84 are not usually able to disperse over large distances (Pérez-Portela & Riesgo, 2018), with 85 some exceptions (see Maldonado, 2006). This limited dispersal capabilities generally result in 86 highly structured and isolated populations (Pérez-Portela & Riesgo, 2018), with high levels of 87 inbreeding and a consequently reduced resilience (Botsford, White, et al., 2009). Hence, to 88 assess the degree of resilience that Antarctic sponges will have under future predicted habitat 89 shifts (IPCC 5th Assessment Report, 2013), it is urgent to investigate their connectivity 90 patterns and gene flow. 4 91 Population genetics, which delves into the distribution of genetic diversity within and 92 between populations, depends essentially on the presence of genetic variability to work with. 93 The mitochondrial genome (mitogenome, mtDNA) has been widely used for population 94 genetic and phylogenetic analyses in Metazoa (Avise, Arnold, et al., 1987) due to its high 95 substitution rates (Brown, George, & Wilson, 1979) and its maternal inheritance and haploidy 96 (see Ernster & Schatz, 1981). However, in some early-splitting animal lineages, such as the 97 members of the phylum Porifera, mtDNA variation within and between species is extremely 98 low, due to its slow-evolving nature (Huang, Meier, Todd, & Chou, 2008a). With some 99 notable exceptions (Duran & Rützler, 2006; DeBiasse, Richards, & Shivji, 2010; López- 100 Legentil & Pawlik, 2009; Xavier, Rachello-Dolmen, et al., 2010), intraspecific relationships 101 in sponges have therefore only been recently addressed using microsatellites (e.g. Calderón, 102 Ortega, et al., 2007; Blanquer & Uriz, 2010; Giles, Saenz-Agudelo, et al., 2015; Riesgo, 103 Pérez-Portela, et al., 2016; Taboada, Riesgo, et al., 2018). Within the past few years, new 104 promising approaches for population genetics based on reduced representation genomic 105 libraries combined with high-throughput sequencing techniques, like restriction-associated 106 DNA sequencing (RADseq) and genotyping by sequencing (GBS), have become routinely 107 implemented in marine invertebrates but hardly on early-splitting lineages (reviewed in Pérez- 108 Portela & Riesgo, 2018). These methods are revolutionizing the ecological, evolutionary and 109 conservation genetic fields because of their power to recover hundreds to thousands of neutral 110 single nucleotide polymorphisms (SNPs) for fine-scale population analyses (Andrews, Good, 111 et al., 2016). However, only one study to date recovered SNPs in sponges, which used 112 amplicon sequencing to obtain 67 SNPs and detect the small-scale genetic structure of 113 Aphrocallistes vastus (Brown, Davis, & Leys, 2017), the main reef-building glass sponge of 5 114 the British Columbia continental shelf. 115 To date, the analysis of RADseq-derived SNPs has just reached Antarctic